Clinical Genomic and Proteomic Approaches to Biomarker Discovery

---

Early detection is one of the most feasible and realistic approaches to improving cancer outcome. The goal of this project is the development of simple blood tests that can be applied to cancer detection. However, the protein composition of human plasma/serum samples varies significantly from individual to individual, and accidental identification of proteins that represent mere personal heterogeneity as biomarker candidates should be avoided by comparing a sufficient number of cases and controls. As a member team in the International Cancer Biomarker Consortium (ICBC) we have been collecting serum/plasma samples from 7 medical institutions participating in the "Third-Term Comprehensive Control Research for Cancer" conducted by the Ministry of Health, Labor, and Welfare, Japan.


Biomarker discovery by direct profiling of plasma/serum proteins
By using high-resolution matrix-assisted laser desorption/ionization hybrid quadrupole time of flight mass spectrometry (MALDI-QqTOF-MS) (1) and original MS data quantification software, namely NCC-ProteoJudge, we were able to directly quantify the entire plasma/serum proteins (or proteome) and succeeded in identifying biomarkers that may be useful for the early diagnosis of pancreatic cancer (2), renal cell cancer (3), and uterine endometrial cancer (4). We recently accomplished a large-scale multi-institutional validation study, in which 8 medical institutions in Japan and Germany participated, and confirmed the significance of biomarkers for early detection of pancreatic cancer.

Possible prediction of radiochemosensitivity of cancer by proteomics
Establishment of a reliable method of predicting the efficacy of chemotherapy and radiotherapy is necessary to provide the most suitable treatment for each cancer patient. MS spectra were obtained from a training set of 27 serum samples (15 pathologically diagnosed responders to PCRT and 12 non-responders). A prediction model was built by machine learning. This set of MS peaks, i.e., the classifier, correctly diagnosed chemoradiosensitivity in 93.3% (14/15) of the cases in the independent validation set in a blinded manner (1).

Development of a New Shotgun Proteomic Platform: 2DICAL
Nano-flow HPLC (high performance liquid chromatography) separation coupled with high-speed MS (mass spectromery) scanning has recently attracted considerable attention because of its comprehensive protein identification capacity, but it has not been recognized as a tool for quantitative proteomics. By eliminating all procedures that reduced the sensitivity and reproducibility of nanoLC-MS, such as isotope labeling, multidimensional LC-separation, and MS/MS (tandem mass spectrometry), we developed an integrated proteome platform, namely 2-Dimensional Image Converted Analysis of Liquid chromatography and mass spectrometry (2DICAL(PDF:293KB)) (6)(7). This high-throughput platform was shown to be successfully extended to quantitative proteomic analysis of trace amounts of formalin-fixed paraffin-embedded (FFPE) tissue samples obtained by lazer micro-dissection(PDF:238KB) (8), assuring mining of FFPE archival samples with detailed clinicopathological records in search of more reliable biomarkers. With use of 2DICAL, we recently identified a biomarker to predict risk of hematologic adverse events (AEs) in pancreatic cancer patients who received gemocitabine monotherapy(PDF:63KB) (9). Based on our finding, we developed a nomogram to predict the risk, which would help to identify who should not receive gemocitabine monotherapy (9).

Distinct gene-expression-defined classes of gastrointestinal stromal tumor
Establishment of a reliable method of predicting recurrence is necessary to provide the most suitable treatment for each cancer patient. We analyzed a well characterized cohort of gastrointestinal stromal tumor (GIST) cases in order to clarify the genomic basis behind the malignant progression of this tumor and to identify a biomarker that might be applicable to the prediction of outcome in GIST patients (10). The expression of dipeptidyl peptidase IV (DPP4/T-cell activation antigen CD26) protein was significantly associated with poorer overall and disease-free survival of gastric GIST(PDF:25KB) (P<.00001).


References
1. Hayashida et al., Clin Cancer Res 11:8042-8047, 2005.
2. Honda et al., Cancer Res 65:10613-10622, 2005.
3. Hara et al., J Urol 174:1213-1217, 2005.
4. Kikuchi et al., Cancer Sci 98:822-829, 2007.
5. Honda et al., Nippon Rinsho 64:1745-55, 2006.
6. Ono et al., Mol Cell Proteomics 5:1338-47, 2006.
7. Negishi et al., Cancer Sci 100:1605-11, 2009
8. Ono et al., J Biol Chem (in press)
9. Matsubara et al., J Clin Oncol 27:2261-8, 2009
10. Yamaguchi et al., J Clin Oncol 26:4100-8, 2008